Crosby.] * Q [October 4, 



The question arises, then, where are these fractures ? A posi- 

 tive answer must be found, because their existence cannot be 

 denied. They are not the planes of division which we call slaty 

 cleavage, because earthquakes are not limited to a particular class 

 of rocks, and successive shocks do not run always in the same 

 direction. They are not the planes of weakness coinciding with 

 the stratification of stratified rocks, for the same and other rea- 

 sons. So far as I can see, we have but one alternative, and that 

 is to call them joints. They are plainly not contraction or expan- 

 sion joints, and hence by a process of elimination we arrive at the 

 conclusion that earthquakes are a principal cause, at least, of 

 ordinary or parallel joints. 



We have already noticed the more important characteristics of 

 this class of joints ; and I invite attention now to a consideration 

 of the probable nature of earthquake-fractures, with a view to 

 comparing them with parallel joints. During the passage of an 

 earthquake, the rocks are subjected in rapid alternation to power- 

 ful compression and equally powerful rarefaction or tension, the 

 latter, probably, being the chief cause of the fractures. The 

 main point is that the fractures must be parallel to the wave-sur- 

 face, i. e., normal at every point to the radius of the sphere or the 

 direction of the wave-movement. This is clearly observed in the 

 case of fractured buildings, which are traversed by the wave in 

 precisely the same manner as the rocks on which they stand ; and 

 nearly all determinations of the position of the seismic vertical 

 and focus of earthquakes rest upon this assumption. 



It follows from this that the fractures will be horizontal between 

 the focus and epicentrum, and that they will dip away from the 

 epicentrum in all directions as we recede from that point, the 

 inclination to the horizon steadily increasing with the distance. 

 And since the focus, according to our best determinations, is 

 rarely more than ten, and often considerably less than five miles 

 below the surface, the fractures must, as a rule, be approximately 

 vertical at all points beyond a few miles from the epicentrum. 

 Now joint-plares ; although sometimes nearly or quite horizontal, 

 are usually approximately vertical, showing a close agreement in 

 this respect with earthquake fractures. 



The point has been raised that the distance from the epicen- 

 trum at which earthquake fractures would become as nearly 



